--------------------------------------------------------------------------------
       log:  K:\Projects\IGO\IGOTables.log
  log type:  text
 opened on:  24 Apr 2005, 01:18:38

. do "C:\DOCUME~1\ALEXMO~1\LOCALS~1\Temp\STD01000000.tmp"

. /*Load data file first*/
. 
. /*TABLE 1 RESULTS*/
. 
. /*Replication: Table 1, Column 1*/
. /*Table 1, Column 2 of Oneal and Russett 1999 with hegdefb and disp_spl and IG
> OSame*/*/
. #del ;
delimiter now ;
. logit disp_l1 
> IGOSame
> smldmat smldep 
> lcaprat2 allies hegdefb
> contig logdstab majpower
> disp_spl*,
> cluster(dyadid) nolog;

Logit estimates                                   Number of obs   =     149403
                                                  Wald chi2(13)   =    2256.35
                                                  Prob > chi2     =     0.0000
Log pseudolikelihood = -5547.8486                 Pseudo R2       =     0.3066

                           (standard errors adjusted for clustering on dyadid)
------------------------------------------------------------------------------
             |               Robust
     disp_l1 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     IGOSame |   .0127644   .0050787     2.51   0.012     .0028103    .0227185
     smldmat |  -.0635548   .0097168    -6.54   0.000    -.0825993   -.0445103
      smldep |  -42.97553   12.25453    -3.51   0.000    -66.99396    -18.9571
    lcaprat2 |  -.2038683   .0418849    -4.87   0.000    -.2859613   -.1217754
      allies |  -.3255853    .156208    -2.08   0.037    -.6317473   -.0194234
     hegdefb |   7.655118     1.7488     4.38   0.000     4.227533     11.0827
    contigkb |   1.767071   .1544214    11.44   0.000     1.464411    2.069732
    logdstab |  -.4501535   .0529521    -8.50   0.000    -.5539377   -.3463693
    majpower |   1.944225   .1467958    13.24   0.000      1.65651    2.231939
   disp_spl0 |  -.6200598   .0338504   -18.32   0.000    -.6864054   -.5537142
   disp_spl1 |   -.006456   .0004331   -14.91   0.000    -.0073048   -.0056072
   disp_spl2 |   .0032939   .0002436    13.52   0.000     .0028165    .0037712
   disp_spl3 |  -.0002518   .0000328    -7.68   0.000    -.0003161   -.0001876
       _cons |  -.9621596   .4327672    -2.22   0.026    -1.810368   -.1139515
------------------------------------------------------------------------------

. /*Our base model: Table 1, Column 2*/
> #del ;
delimiter now ;
. logit disp_l1
> IGOSame ClusSame CentDif ClusSizeMax 
> smldmat smldep
> lcaprat2 allies hegdefb
> contig logdstab majpower 
> disp_spl*,
> cluster(dyadid) nolog;

Logit estimates                                   Number of obs   =     149403
                                                  Wald chi2(16)   =    2190.46
                                                  Prob > chi2     =     0.0000
Log pseudolikelihood =  -5527.218                 Pseudo R2       =     0.3092

                           (standard errors adjusted for clustering on dyadid)
------------------------------------------------------------------------------
             |               Robust
     disp_l1 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     IGOSame |   .0151643   .0051661     2.94   0.003     .0050388    .0252898
    ClusSame |  -.1605983   .1005341    -1.60   0.110    -.3576415    .0364449
     CentDif |  -.0003548   .0001031    -3.44   0.001    -.0005569   -.0001528
 ClusSizeMax |   .0116839    .003798     3.08   0.002       .00424    .0191277
     smldmat |  -.0652762   .0093929    -6.95   0.000     -.083686   -.0468664
      smldep |  -41.90489   12.54339    -3.34   0.001    -66.48949   -17.32029
    lcaprat2 |  -.1891118   .0437274    -4.32   0.000     -.274816   -.1034076
      allies |  -.3660362    .159569    -2.29   0.022    -.6787857   -.0532866
     hegdefb |   7.912999   1.730503     4.57   0.000     4.521276    11.30472
    contigkb |   1.743877   .1567875    11.12   0.000     1.436579    2.051175
    logdstab |   -.456574   .0565603    -8.07   0.000    -.5674302   -.3457178
    majpower |   1.997726   .1525801    13.09   0.000     1.698675    2.296778
   disp_spl0 |  -.6229316   .0339528   -18.35   0.000    -.6894779   -.5563852
   disp_spl1 |  -.0064649   .0004337   -14.91   0.000    -.0073149   -.0056149
   disp_spl2 |   .0032965   .0002438    13.52   0.000     .0028186    .0037744
   disp_spl3 |  -.0002517   .0000328    -7.67   0.000     -.000316   -.0001874
       _cons |  -1.108581   .4702065    -2.36   0.018    -2.030169   -.1869936
------------------------------------------------------------------------------

. /*other specifications in order to check robustness*/
> 
> /*MinVarModel: Table 1, Column 3*/
> #del ;
delimiter now ;
. logit disp_l1 
> IGOSame ClusSame CentDif ClusSizeMax
> contig logdstab majpower
> disp_spl*,
> cluster(dyadid) nolog;

Logit estimates                                   Number of obs   =     149403
                                                  Wald chi2(11)   =    1718.19
                                                  Prob > chi2     =     0.0000
Log pseudolikelihood = -5711.4213                 Pseudo R2       =     0.2862

                           (standard errors adjusted for clustering on dyadid)
------------------------------------------------------------------------------
             |               Robust
     disp_l1 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     IGOSame |   .0103754   .0040391     2.57   0.010     .0024588     .018292
    ClusSame |   -.341698   .1062691    -3.22   0.001    -.5499815   -.1334145
     CentDif |  -.0002441   .0000888    -2.75   0.006    -.0004182     -.00007
 ClusSizeMax |   .0176424   .0040051     4.40   0.000     .0097925    .0254923
    contigkb |   1.741324    .170321    10.22   0.000     1.407501    2.075147
    logdstab |  -.3760356     .05932    -6.34   0.000    -.4923006   -.2597707
    majpower |   1.514058   .1466218    10.33   0.000     1.226684    1.801431
   disp_spl0 |  -.6349878   .0339411   -18.71   0.000    -.7015112   -.5684645
   disp_spl1 |  -.0065536   .0004314   -15.19   0.000    -.0073992    -.005708
   disp_spl2 |   .0033424   .0002419    13.82   0.000     .0028683    .0038165
   disp_spl3 |  -.0002558    .000032    -7.98   0.000    -.0003186    -.000193
       _cons |  -1.331081   .4820905    -2.76   0.006    -2.275961   -.3862013
------------------------------------------------------------------------------

. /*PR Dyads: Table 1, Column 4*/
> #del ;
delimiter now ;
. logit disp_l1 
> IGOSame ClusSame CentDif ClusSizeMax
> smldmat smldep 
> lcaprat2 allies hegdefb 
> contig logdstab majpower 
> disp_spl*
> if majpower==1 | contig==1,
> cluster(dyadid) nolog;

Logit estimates                                   Number of obs   =      33354
                                                  Wald chi2(16)   =    1141.16
                                                  Prob > chi2     =     0.0000
Log pseudolikelihood = -4077.9977                 Pseudo R2       =     0.2290

                           (standard errors adjusted for clustering on dyadid)
------------------------------------------------------------------------------
             |               Robust
     disp_l1 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     IGOSame |   .0101115   .0047341     2.14   0.033     .0008329    .0193901
    ClusSame |   -.247072   .1015386    -2.43   0.015    -.4460839     -.04806
     CentDif |  -.0004475   .0001114    -4.02   0.000     -.000666   -.0002291
 ClusSizeMax |    .007091   .0038504     1.84   0.066    -.0004556    .0146377
     smldmat |  -.0606138   .0091088    -6.65   0.000    -.0784668   -.0427608
      smldep |  -21.70292   9.615182    -2.26   0.024    -40.54833   -2.857513
    lcaprat2 |  -.2559083   .0441651    -5.79   0.000    -.3424703   -.1693463
      allies |  -.3271722   .1548494    -2.11   0.035    -.6306715   -.0236729
     hegdefb |   8.347307   1.780026     4.69   0.000     4.858521    11.83609
    contigkb |   .8111283    .140411     5.78   0.000     .5359278    1.086329
    logdstab |  -.2146992   .0545758    -3.93   0.000    -.3216658   -.1077327
    majpower |    .793507   .1595403     4.97   0.000     .4808137      1.1062
   disp_spl0 |  -.5968805    .034713   -17.19   0.000    -.6649167   -.5288443
   disp_spl1 |  -.0059371   .0004563   -13.01   0.000    -.0068315   -.0050428
   disp_spl2 |   .0029869   .0002589    11.54   0.000     .0024796    .0034943
   disp_spl3 |  -.0002109   .0000355    -5.94   0.000    -.0002806   -.0001413
       _cons |  -.8012736   .4415326    -1.81   0.070    -1.666662    .0641144
------------------------------------------------------------------------------

. /*Dispute Onset: Table 1, Column 5*/
> #del ;
delimiter now ;
. logit dispon_l1
> IGOSame ClusSame CentDif ClusSizeMax
> smldmat smldep 
> lcaprat2 allies hegdefb 
> contig logdstab majpower
> dispon_spl*,
> cluster(dyadid) nolog;

Logit estimates                                   Number of obs   =     149403
                                                  Wald chi2(16)   =    1915.63
                                                  Prob > chi2     =     0.0000
Log pseudolikelihood = -4264.0754                 Pseudo R2       =     0.1751

                           (standard errors adjusted for clustering on dyadid)
------------------------------------------------------------------------------
             |               Robust
   dispon_l1 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     IGOSame |  -.0112763   .0051457    -2.19   0.028    -.0213618   -.0011909
    ClusSame |  -.0516875   .0997563    -0.52   0.604    -.2472063    .1438312
     CentDif |  -.0003418   .0001043    -3.28   0.001    -.0005462   -.0001374
 ClusSizeMax |   .0113965    .003984     2.86   0.004     .0035879    .0192051
     smldmat |  -.0531216   .0084168    -6.31   0.000    -.0696182    -.036625
      smldep |   -26.2047   11.05337    -2.37   0.018    -47.86891   -4.540496
    lcaprat2 |  -.1819541   .0389158    -4.68   0.000    -.2582277   -.1056805
      allies |  -.2772944   .1364797    -2.03   0.042    -.5447897   -.0097991
     hegdefb |    9.67049   1.504241     6.43   0.000     6.722232    12.61875
    contigkb |   1.697875   .1411856    12.03   0.000     1.421156    1.974593
    logdstab |  -.5047074   .0473183   -10.67   0.000    -.5974496   -.4119653
    majpower |   1.763725   .1359518    12.97   0.000     1.497264    2.030186
 dispon_spl0 |  -.1051933   .0285241    -3.69   0.000    -.1610995   -.0492871
 dispon_spl1 |  -.0012982   .0004018    -3.23   0.001    -.0020857   -.0005108
 dispon_spl2 |   .0006875   .0002335     2.94   0.003     .0002299    .0011452
 dispon_spl3 |  -.0000601   .0000338    -1.78   0.076    -.0001264    6.22e-06
       _cons |  -2.270208   .4299931    -5.28   0.000    -3.112979   -1.427437
------------------------------------------------------------------------------

. /*GEE: Table 1, Column 6*/
> #del ;
delimiter now ;
. set matsize 800;

. xtgee disp_l1 
> IGOSame ClusSame CentDif ClusSizeMax 
> smldmat smldep 
> lcaprat2 allies hegdefb
> contig logdstab majpower, 
> family(binomial) link(logit) corr(ar1) force robust nolog;
note:  some groups have fewer than 2 observations
       not possible to estimate correlations for those groups
       31 groups omitted from estimation


GEE population-averaged model                   Number of obs      =    149372
Group and time vars:           dyadid year      Number of groups   =      5829
Link:                                logit      Obs per group: min =         2
Family:                           binomial                     avg =      25.6
Correlation:                         AR(1)                     max =        90
                                                Wald chi2(12)      =   1386.87
Scale parameter:                         1      Prob > chi2        =    0.0000

                           (standard errors adjusted for clustering on dyadid)
------------------------------------------------------------------------------
             |             Semi-robust
     disp_l1 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     IGOSame |  -.0124633   .0052797    -2.36   0.018    -.0228113   -.0021152
    ClusSame |  -.1250222    .095322    -1.31   0.190      -.31185    .0618056
     CentDif |  -.0002975   .0001208    -2.46   0.014    -.0005342   -.0000608
 ClusSizeMax |   .0081983   .0033631     2.44   0.015     .0016068    .0147898
     smldmat |  -.0656726   .0101926    -6.44   0.000    -.0856497   -.0456955
      smldep |  -46.78243   15.03568    -3.11   0.002    -76.25183   -17.31303
    lcaprat2 |  -.2610732   .0537626    -4.86   0.000    -.3664458   -.1557005
      allies |  -.3728577   .1703989    -2.19   0.029    -.7068333    -.038882
     hegdefb |   13.45434   1.691431     7.95   0.000     10.13919    16.76948
    contigkb |   1.936558   .1874226    10.33   0.000     1.569217      2.3039
    logdstab |   -.537004   .0593933    -9.04   0.000    -.6534127   -.4205953
    majpower |   2.021985   .1840225    10.99   0.000     1.661308    2.382663
       _cons |  -1.932543    .491652    -3.93   0.000    -2.896163   -.9689229
------------------------------------------------------------------------------

. /*TABLE 2 PROBABILITIES*/
> /*Our base model:*/
> #del ;
delimiter now ;
. logit disp_l1 
> IGOSame ClusSame CentDif ClusSizeMax
> smldmat smldep
> lcaprat2 allies hegdefb
> contig logdstab majpower 
> disp_spl*,
> cluster(dyadid) nolog;

Logit estimates                                   Number of obs   =     149403
                                                  Wald chi2(16)   =    2190.46
                                                  Prob > chi2     =     0.0000
Log pseudolikelihood =  -5527.218                 Pseudo R2       =     0.3092

                           (standard errors adjusted for clustering on dyadid)
------------------------------------------------------------------------------
             |               Robust
     disp_l1 |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     IGOSame |   .0151643   .0051661     2.94   0.003     .0050388    .0252898
    ClusSame |  -.1605983   .1005341    -1.60   0.110    -.3576415    .0364449
     CentDif |  -.0003548   .0001031    -3.44   0.001    -.0005569   -.0001528
 ClusSizeMax |   .0116839    .003798     3.08   0.002       .00424    .0191277
     smldmat |  -.0652762   .0093929    -6.95   0.000     -.083686   -.0468664
      smldep |  -41.90489   12.54339    -3.34   0.001    -66.48949   -17.32029
    lcaprat2 |  -.1891118   .0437274    -4.32   0.000     -.274816   -.1034076
      allies |  -.3660362    .159569    -2.29   0.022    -.6787857   -.0532866
     hegdefb |   7.912999   1.730503     4.57   0.000     4.521276    11.30472
    contigkb |   1.743877   .1567875    11.12   0.000     1.436579    2.051175
    logdstab |   -.456574   .0565603    -8.07   0.000    -.5674302   -.3457178
    majpower |   1.997726   .1525801    13.09   0.000     1.698675    2.296778
   disp_spl0 |  -.6229316   .0339528   -18.35   0.000    -.6894779   -.5563852
   disp_spl1 |  -.0064649   .0004337   -14.91   0.000    -.0073149   -.0056149
   disp_spl2 |   .0032965   .0002438    13.52   0.000     .0028186    .0037744
   disp_spl3 |  -.0002517   .0000328    -7.67   0.000     -.000316   -.0001874
       _cons |  -1.108581   .4702065    -2.36   0.018    -2.030169   -.1869936
------------------------------------------------------------------------------

. #del ;
delimiter now ;
. /*Baseline*/
> prvalue,  rest(mean) brief;

  Pr(y=1|x):          0.0024   95% ci: (0.0021,0.0027)
  Pr(y=0|x):          0.9976   95% ci: (0.9973,0.9979)

. /*cluster*/
> prvalue, x(ClusSame=min) rest(mean) brief;

  Pr(y=1|x):          0.0025   95% ci: (0.0021,0.0028)
  Pr(y=0|x):          0.9975   95% ci: (0.9972,0.9979)

. prvalue, x(ClusSame=max) rest(mean) brief;

  Pr(y=1|x):          0.0021   95% ci: (0.0017,0.0026)
  Pr(y=0|x):          0.9979   95% ci: (0.9974,0.9983)

. /*Prestige*/
> prvalue, x(CentDif=min) rest(mean);
 
logit: Predictions for disp_l1

  Pr(y=1|x):          0.0029   95% ci: (0.0025,0.0035)
  Pr(y=0|x):          0.9971   95% ci: (0.9965,0.9975)

        IGOSame     ClusSame      CentDif  ClusSizeMax      smldmat
x=    22.297517    .27168798            0    30.404115   -3.3194113

         smldep     lcaprat2       allies      hegdefb     contigkb
x=    .00118759    1.9894589    .13347791    .05776479    .08061418

       logdstab     majpower    disp_spl0    disp_spl1    disp_spl2
x=    8.0396073    .17556542    21.502379   -11524.163   -21499.101

      disp_spl3
x=   -29319.949

. prvalue, x(CentDif=max) rest(mean);
 
logit: Predictions for disp_l1

  Pr(y=1|x):          0.0006   95% ci: (0.0003,0.0014)
  Pr(y=0|x):          0.9994   95% ci: (0.9986,0.9997)

        IGOSame     ClusSame      CentDif  ClusSizeMax      smldmat
x=    22.297517    .27168798         4343    30.404115   -3.3194113

         smldep     lcaprat2       allies      hegdefb     contigkb
x=    .00118759    1.9894589    .13347791    .05776479    .08061418

       logdstab     majpower    disp_spl0    disp_spl1    disp_spl2
x=    8.0396073    .17556542    21.502379   -11524.163   -21499.101

      disp_spl3
x=   -29319.949

. /*Size*/
> prvalue, x(ClusSizeMax=min) rest(mean);
 
logit: Predictions for disp_l1

  Pr(y=1|x):          0.0017   95% ci: (0.0013,0.0022)
  Pr(y=0|x):          0.9983   95% ci: (0.9978,0.9987)

        IGOSame     ClusSame      CentDif  ClusSizeMax      smldmat
x=    22.297517    .27168798    632.71152            3   -3.3194113

         smldep     lcaprat2       allies      hegdefb     contigkb
x=    .00118759    1.9894589    .13347791    .05776479    .08061418

       logdstab     majpower    disp_spl0    disp_spl1    disp_spl2
x=    8.0396073    .17556542    21.502379   -11524.163   -21499.101

      disp_spl3
x=   -29319.949

. prvalue, x(ClusSizeMax=max) rest(mean);
 
logit: Predictions for disp_l1

  Pr(y=1|x):          0.0030   95% ci: (0.0025,0.0037)
  Pr(y=0|x):          0.9970   95% ci: (0.9963,0.9975)

        IGOSame     ClusSame      CentDif  ClusSizeMax      smldmat
x=    22.297517    .27168798    632.71152           52   -3.3194113

         smldep     lcaprat2       allies      hegdefb     contigkb
x=    .00118759    1.9894589    .13347791    .05776479    .08061418

       logdstab     majpower    disp_spl0    disp_spl1    disp_spl2
x=    8.0396073    .17556542    21.502379   -11524.163   -21499.101

      disp_spl3
x=   -29319.949

. /*Social Allies/Rivals*/
> prvalue, x(ClusSame=min CentDif=min ClusSizeMax=max) rest(mean) brief;

  Pr(y=1|x):          0.0040   95% ci: (0.0032,0.0049)
  Pr(y=0|x):          0.9960   95% ci: (0.9951,0.9968)

. prvalue, x(ClusSame=max CentDif=max ClusSizeMax=min) rest(mean) brief;

  Pr(y=1|x):          0.0004   95% ci: (0.0002,0.0009)
  Pr(y=0|x):          0.9996   95% ci: (0.9991,0.9998)

. /*Dem*/
> prvalue, x(smldmat=min) rest(mean) brief;

  Pr(y=1|x):          0.0036   95% ci: (0.0032,0.0042)
  Pr(y=0|x):          0.9964   95% ci: (0.9958,0.9968)

. prvalue, x(smldmat=max) rest(mean) brief;

  Pr(y=1|x):          0.0010   95% ci: (0.0007,0.0014)
  Pr(y=0|x):          0.9990   95% ci: (0.9986,0.9993)

. /*Dep*/
> prvalue, x(smldep=min) rest(mean);
 
logit: Predictions for disp_l1

  Pr(y=1|x):          0.0025   95% ci: (0.0022,0.0028)
  Pr(y=0|x):          0.9975   95% ci: (0.9972,0.9978)

        IGOSame     ClusSame      CentDif  ClusSizeMax      smldmat
x=    22.297517    .27168798    632.71152    30.404115   -3.3194113

         smldep     lcaprat2       allies      hegdefb     contigkb
x=            0    1.9894589    .13347791    .05776479    .08061418

       logdstab     majpower    disp_spl0    disp_spl1    disp_spl2
x=    8.0396073    .17556542    21.502379   -11524.163   -21499.101

      disp_spl3
x=   -29319.949

. prvalue, x(smldep=max) rest(mean);
 
logit: Predictions for disp_l1

  Pr(y=1|x):          0.0000   95% ci: (0.0000,0.0001)
  Pr(y=0|x):          1.0000   95% ci: (0.9999,1.0000)

        IGOSame     ClusSame      CentDif  ClusSizeMax      smldmat
x=    22.297517    .27168798    632.71152    30.404115   -3.3194113

         smldep     lcaprat2       allies      hegdefb     contigkb
x=    .20846155    1.9894589    .13347791    .05776479    .08061418

       logdstab     majpower    disp_spl0    disp_spl1    disp_spl2
x=    8.0396073    .17556542    21.502379   -11524.163   -21499.101

      disp_spl3
x=   -29319.949

. 
end of do-file

. log close
       log:  K:\Projects\IGO\IGOTables.log
  log type:  text
 closed on:  24 Apr 2005, 01:26:07
--------------------------------------------------------------------------------
